Organometallic cyclometallated complexes of transition metals (e.g. rhodium, iridium, platinum) have become useful materials because of their photophysical and photochemical properties. One especially important application of these compounds are as phosphorescent dopants in Organic Light-Emitting Diodes (OLEDs) because of their strong emission from triplet excited states (M. A. Baldo, et al, Appl. Phys. Letters, 75, 4 (1999)). An important class of phosphorescent cyclometallated complexes contain ligands that are at least bidentate wherein one coordination site of the ligand to the metal is through an N atom that is doubly bonded to C or another N atom, usually as part of a heterocyclic ring, and wherein another coordination site of the ligand to the metal is through a C atom. As used herein, the term “organometallic cyclometallated complex” means that at least one of the coordination sites forming the cyclic unit binding the metal atom by at least one ligand must be a metal-carbon bond. The metal-carbon bond is formed in place of a hydrogen-carbon bond of the free ligand before it is complexed. The carbon atom forming the metal carbon bond is usually also doubly bonded to another carbon as in, for example, a phenyl ring or a thienyl ring or furanyl ring. Further the carbon atom forming the metal-carbon bond also is preferably positioned so as to form a five- or six-membered metallacycle including the coordinated N atom of the ligand. A tris-cyclometallated complex has three such ligands. Some examples of iridium(III) organometallic cyclometallated complexes are shown below. It is also possible that the organometallic cyclometallating ligands are not all the same.

Further, there are two isomers, facial and meridional (fac and mer), possible for such complexes having three identical but unsymmetrical bidentate ligands as illustrated below. The facial isomers are typically more desirable in OLED applications because they usually have higher quantum efficiencies.

The usefulness and importance of organometallic cyclometallated complexes of second- and third-row transition metals have necessitated synthetic methods for preparing them more efficiently. Chassot et al., Inorg. Chem., 23, 4249-4253, (1984), have used lithiated ligands with platinum compounds that include leaving groups to form cyclometallated complexes of the ligands with platinum. Jolliet et al., Inorg. Chem., 35, 4883-4888, (1996) also used lithiated ligands to form cyclometallated complexes of the ligands with platinum or palladium, and Lamansky and Thompson, in WO 00/57676, used the same procedure for cyclometallated platinum complexes. These procedures suffer from low yields, as well as the relative instability of and difficulty in handling lithiated organic materials.
Organometallic cyclometallated complexes may also be formed from direct reaction of the cyclometallating ligand, wherein the carbon-hydrogen is activated and replaced by the carbon-metal bond. For example, fac-tris(2-phenylpyridinato-N,C2′)iridium(III), or Ir(ppy)3, was made by reaction of 2-phenylpyridine and tris(acetylacetonate) iridium (Ir(acac)3) in glycerol solvent by K. Dedian et al., Inorg. Chem., 30, 1685 (1991). Stössel and coworkers (WO02060910) further optimized and improved this reaction, but still using the expensive Ir(acac)3 starting material. By reacting less expensive halide complexes of Ir(III) such as iridium(III) chloride hydrate with 2-phenylpyridine in a solvent comprising a 3:1 mixture of 2-ethoxy-ethanol and water, Nonoyama obtained dimeric organometallic cyclometallated complexes such as tetrakis(2-phenyl-pyridinato-N,C2′-) (di-μ-chloro)di-iridium(III). (Ir(ppy)3 was later extracted as a side product in 10% yield from this reaction mixture, K. A. King, et al., J. Am. Chem. Soc., 107, 1431 (1985).) M. G. Colombo, et al., Inorg Chem., 33, 545 (1994), further reacted the above-cited di-iridium complex with a silver salt in neat 2-phenylpyridine to obtain Ir(ppy)3 in 75% yield. Grushin et al., US 2002/0190250, used this process to make additional tris-cyclometallated complexes of Ir(III) having fluorine-substitutions on phenylpyridine and phenylquinoline cyclometallating ligands. H. Konno and Y. Sasaki, Chem. Lett., 32, 252 (2003) found that fac-Ir(ppy)3 could be formed in good yield by the reaction of IrCl3 3H2O with a excess of the ligand under microwave conditions. Many of these processes require a large excess of a ligand since it is employed as the solvent, thereby either consuming valuable material or necessitating a process to recover excess ligand.
Lamasky et al., Inorg. Chem., 40, 1704-1711, (2001), demonstrated yet another process for making tris-cyclometallated iridium complexes. First, a mixed ligand complex, bis(7,8-benzoquinolinato-N,C3′)iridium(III)(acetylacetonate), was made from tetrakis(7,8-benzoquinolinato-N,C3′)(di-μ-chloro)di-iridium(III). Then the bis(7,8-benzoquinolinato-N,C3′)iridium(III)(acetylacetonate) was reacted with additional 7,8-benzoquinoline in refluxing glycerol to produce a mixture of isomers of the tris-cyclometallated complex, tris(7,8-benzoquinolinato-N,C3′)iridium(III). Kamatani et al., US 2003/0068526, have also employed this reaction type for additional cyclometallated iridium complexes. But this process often yields less desirable meridional isomers or mixtures of the facial and meridional isomers of the tris-cyclometallated complexes. This process also requires very long reaction times at elevated temperatures, in the case of many other desired ligands, to completely replace the acetylacetonate or similar anionic bidentate ligand with the desired organometallic cyclometallating ligand. Tamayo et al., J. Am. Chem. Soc., 125, 7377-7387 (2003), have shown that reaction of dimeric organometallic cyclometallated complexes such as tetrakis(2-phenyl-pyridinato-N,C2′) (di-μ-chloro)di-iridium(III) with sodium carbonate and additional cyclometallating ligand in glycerol can lead to formation of meridional isomers in many cases, while further reaction at higher temperatures results in formation of mostly facial isomer. However, this procedure is inconvenient for facial isomers as it necessitates finding exact conditions for the reaction of each ligand.
U.S. Pat. No. 6,870,054 describes a process for forming organometallic cyclometallated complexes of Ir(III) comprising the step of reacting a halide-containing complex of the metal with a silver salt and a heterocyclic organic ligand compound capable of forming an organometallic cyclometallated complex and in a solvent comprising an organic diol. However, this process fails or works poorly in many cases of desirable ligands. One of the possible reasons for this process not being generally applicable to a wide variety of possible cyclometallating ligands is that the solubility of the intermediate complexes may be too low in these solvents.
B. Schmid, F. Garces, and R. Watts, Inorg. Chem., 33, 9 (1994), describe the preparation of solvento complexes of iridium that additionally comprise cyclometallating ligands. However these materials comprise cationic complexes that are not volatile enough for vapor deposition and therefore are not as useful as tris-cyclometallated complexes for OLED applications.
Commonly assigned U.S. Ser. No. 10/879,657, filed on Jun. 29, 2004, describes a process for forming organometallic tris-cyclometallated complexes of Ir(III) or Rh(III) from a complex comprising an Ir (III) or Rh (III) metal ion and two cyclometallated ligands, two monodentate ligands and a counterion. That process requires the formation of a bis-cyclometallated intermediate, which is difficult for certain ligands.
Nakayama and coworkers, WO 2004/043974, describe a process in which a monovalent iridium dinuclear complex, shown below, wherein A represents a non-conjugated diene compound and X represents a halogen atom, is reacted with a heterocyclic compound capable of forming an organometallic cyclometallated complex.

However, the product of this process is a dimeric organometallic cyclometallated complex such as tetrakis(2-phenyl-pyridinato-N,C2′-) (di-μ-chloro)di-iridium(III), and not the desired tris-cyclometallated complex.
Despite the large number of investigations into the synthetic methodology for cyclometallated organometallic complexes, there remains a need for alternative methods that are applicable to a wide range of cyclometallating ligands, especially cyclometallating ligands containing multiple heteroatoms.